Search

A block-inverse preconditioner (BIP) is proposed to accelerate solving implicit time integration in the context of Newton–Krylov approach used in micromagnetic simulations for solving the Landau–Lifshitz–Gilbert equation. A coefficient matrix is generated and stored for the linear system of Newton method. BIP is formulated by subdividing the coefficient matrix into small blocks and directly inverting them. The cost of preconditioning is low since inverting and multiplying small blocks is fast, which can be further minimized by parallel computing on multicore CPUs or GPUs. The effectiveness, speed, robustness, and scalability of BIP is demonstrated by numerical simulation experiments. Comparisons to incomplete LU preconditioning methods are conducted to demonstrate the effectiveness of BIP.

International audience; Spin-transfer-driven oscillations of a transverse domain wall confined to a perpendicular spin-valve structure are investigated using a one-dimensional model. The stack consists of a polarizer, nonmagnetic spacer, soft free layer, and pinned magnetic layer. It is found that the domain-wall oscillation frequency is a nonmonotonic, highly asymmetrical function of applied electric current, showing a strong dependence on the current direction and the relative strengths of the interfacial and bulk spin-transfer torques. Micromagnetic analysis reveals that the surprising and atypical oscillator response is due to an interplay between the interfacial spin-transfer torque, the bulk spin-transfer torque, the exchange torque, and the damping torque. The underlying physical and material responses are examined, including the important role of the domain-wall twist. The competitions between the involved torques under different operating conditions suggest that the oscillator could serve as a model system to investigate magnetic and spintronic phenomena at the nanoscale. The observed current-dependent twisting of the free-layer magnetization about the axis of precession may further be found interesting for investigations of the interaction between spin-polarized current and chiral spin structures.

Fast coupled micromagnetic-electromagnetic solvers are presented. The solvers allow modeling the non-linear magnetization dynamics in complex magnetic structures accounting for dynamic eddy current interactions. The solvers rely on a concurrently integrating in time of the Landau-Lifshitz-Gilbert equation and time-domain electromagnetic integral equatiuon in the magneto-quasi-static approximation. The solvers are implemented on multi-core CPU and high-performance GPU computer systems.

International audience; Current induced magnetization manipulation in a spin valve structure where the free layer is a magnetic hard/soft composite structure is studied using micromagnetic simulations. In this structure where the hard layers has strong perpendicular magnetic anisotropy, a domain wall can be nucleated in the soft layer due to the spin transfer torque effect. Depending on the magnetic properties of the layers and the current intensity the domain wall can induce the free layer reversal or be pinned by the hard layer. For these non-uniform magnetic configurations both bulk and interface spin transfer torques need to be considered. The potential reduction of the critical current observed in this geometry is of potential technological interest.

The energy barriers are studied for an in-plane magnetized elliptic magnetoresistive random access memory cell using the nudged elastic band method with finite-element micromagnetics. Our calculations show that the transition between coherent magnetization rotation and domain wall mediated thermal reversal appears at an ellipse aspect ratio greater than 2 for widths 60-90 nm. This crossover point is almost independent of the value of the partial perpendicular magnetic anisotropy and has only a negligibly weak dependence of the width and thickness of the cell and only weakly depends on the exchange stiffness for reasonable in-plane Spin-Transfer Torque Magnetic Random Access Memory parameters. We developed an analytical model for the thermal barrier, which agrees with the numerical results.

Magnetic tunnel junction (MTJ) are at the center of the magnetic random access memories (MRAM) technology. Nowadays, the research and development efforts are focusing on perpendicularly magnetized layers in the MTJ, taking advantage of the anisotropy provided by spin-orbit coupling at the interfaces with heavy metals such as Pt or Ta [1]. Recently, interfaces between ferromagnetic and heavy metals are also studied in a more fundamental context, because an antisymmetric interaction might appear at the interface, namely an interfacial Dzyaloshinskii-Moriya interaction (DMI) [2]. We will present a micromagnetic study of the reversal of the free layer magnetization direction by spin transfer torque under different amplitudes of the DMI. In particular, we will evaluate the energy barrier E b or equivalently the thermal stability Δ=E b /(300k B ) and the minimal current density necessary for switching J c0 as a function of the DMI amplitude.

Magnetic nanowires supporting field- and current-driven domain wall motion are envisioned for new methods of information storage and processing. A major obstacle for their practical use is the domainwall velocity, which is traditionally limited due to the Walker breakdown occurring when the forcing field or current reaches a critical threshold value. We show through numerical and analytical modeling that the Walker breakdown limit can be extended or completely eliminated in antiferromagnetically coupled magnetic nanowires. These coupled nanowires allow for giant domain-wall velocities driven by field and/or current via spin transfer torque as compared to conventional nanowires.

International audience; In order to increase the thermal stability of a magnetic random access memory cell, materials with high spin-orbit interaction are often introduced in the storage layer. As a side effect, a strong Dzyaloshinskii-Moriya interaction (DMI) may arise in such systems. Here, we investigate the impact of DMI on the magnetic cell performance, using micromagnetic simulations. We find that DMI strongly promotes non-uniform magnetization states and non-uniform switching modes of the magnetic layer. It appears to be detrimental for both the thermal stability of the cell and its switching current, leading to considerable deterioration of the cell performance even for a moderate DMI amplitude.

The feasibility of using high-performance desktop and embedded mobile computational platforms is presented, including multi-core Intel central processing unit, Nvidia desktop graphics processing units, and Nvidia Jetson TK1 Platform. FastMag finite element method-based micromagnetic simulator is used as a testbed, showing high efficiency on all the platforms. Optimization aspects of improving the performance of the mobile systems are discussed. The high performance, low cost, low power consumption, and rapid performance increase of the embedded mobile systems make them a promising candidate for micromagnetic simulations. Such architectures can be used as standalone systems or can be built as low-power computing clusters.